With the development of computer technologies, types of APPs continuously increase, and more and more APPs are installed on a terminal. How to display APPs so as to make it more convenient for a user to search for a desired APP becomes a key technology.
In the prior art, sensor data of a terminal and external environment data are collected; a context environment in which the terminal is located is analyzed by using historical empirical data; APPs that may be used by a user are predicted according to the context environment in which the terminal is located; and the predicted APPs are sorted according to probabilities and displayed on a screen, so as to facilitate selection of a desired APP by the user.
However, according to a method in the prior art, a large amount of sensor data of a terminal and external data need to be collected. Calculation is complex and an amount of calculation is large, which degrades processing performance of the terminal.